Analysts' recommendations and rigorous valuation models

In my Financial Statement Analysis and Valuation class we talk about how the "residual income model" is a rigorous approach to valuing a company and possibly spotting mispriced stocks. Specifically, we can compare the valuation model estimate we come up with using our forecasts and compare them to price (see my earlier post on value-to-price here). This would lead us to being able to recommend whether we believed that the stock was currently under- or overvalued... just as sell-side analysts do!

A curious result in the academic literature was presented by Mark Bradshaw who finds that analysts' don't seem to use the residual income model in generating their stock recommendations but instead appear to use their estimates of expected growth. Now this may seem reasonable until we notice that trading on expected growth would have us lose money, while trading on the residual income model would make us a nice return!

In a paper that I coauthored with Andreas Simon, we were interested in whether this finding was true of the very best analysts, so we investigated whether analysts who are good at forecasting appear to use the residual income model in generating their stock recommendations. We find that those analysts who are "putting in the effort" to produce better forecasts are also more likely to be generating their recommendations from a model "like" the residual income model and have more profitable recommendations (note I say "like" because even this group of analysts appear to be aligning their recommendations with their growth expectations than would be suggested by theory). Our paper can be found at ssrn and has been accepted for publication at the Journal of Business Finance and Accounting.

Valuing facebook

Related to one of our class exercises is the question of how much can these social networking sites be worth?

As a starting point you will notice that all three of the commentators in this article "How much is facebook really worth?" are basing their estimates on multiples of revenue. Of course revenue growth without growth in earnings (and hopefully even residual earnings) tends not to add value in the long-run.

Another point of contention arises between the three authors on the selection of a suitable comparable firm, is it myspace (1st author)? or is it yahoo (last author)? Clearly the assumptions of the value of facebook are widely varied based on the selection of the comparable firm.

Is facebook worth 15 billion?

Is Apple ripe?

I couldn't help partially stealing the title of this article: "BUY OR SELL - Are Apple shares ripe for buying?" the article is essentially discussing whether the market is too heavily discounting Apple shares following the news of Steve Jobs (Apple's CEO) taking leave for health reasons.

As I have mentioned in class a few times now, market prices in the current market suggest to me that we are currently in a buyer's market.

Let's consider Apple's price at around $83 per share (at the time of writing). That's about a PE (price to earnings) ratio of under 16.5, the lowest PE ratio for Apple in the last 5 years is about 15.9 (note that the forward PE ratio, that is when the earnings are the expected earnings for this year and not last year's, the ratio drops to around 11). With an ROE (return on equity) of about 27 and sales growth in excess of 20% how can we reconcile the apparent strength of the company's financials with their PE ratio?

If we think about our model of value, what does this suggest that the market "feels" about the earnings of Apple? For example, does it seem like the pricing is reflecting a low sustainability (or persistence) of this past performance?

Does it appear that the sustainability of their earnings is low?

Does one individual make that much of a difference to the strategy of the company?

Or maybe, it's being undervalued due to speculation...?

Value-to-price

I will post some results from relevant accounting studies over the course of the semester. The first I'd like to highlight relates to today's discussion of the residual income model. The paper I want to highlight examines the predictive ability of the residual income value to price ratio over future stock returns. The paper is by Rich Frankel and Charles Lee, who published "Accounting valuation, market expectation, and cross-sectional stock returns" in the Journal of Accounting Research in 1998.


In their paper they find that over holding periods of 36 months, a strategy based on buying high value to price stocks (i.e., stocks the market undervalues relative to the model of fundamental value) and selling stocks low value to price stocks (i.e., stocks the market overvalues relative to the model of fundamental value) earns over a 45% return.

Do you think that these results are likely to hold in our current economy?